[
https://issues.apache.org/jira/browse/CALCITE-2846?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16819780#comment-16819780
]
Forward Xu commented on CALCITE-2846:
-------------------------------------
Reference blink now is to write its own FlinkSqlOperatorTable class, which
function to use, directly from StandardXX OracleXXX reference. That's pretty
clear. In fact, this practice links the function chains of different databases
together.
> Document Oracle-specific functions, such as NVL and LTRIM, in the SQL
> reference
> -------------------------------------------------------------------------------
>
> Key: CALCITE-2846
> URL: https://issues.apache.org/jira/browse/CALCITE-2846
> Project: Calcite
> Issue Type: Improvement
> Components: site
> Reporter: Julian Hyde
> Priority: Major
> Labels: documentation
>
> Document Oracle-specific functions (DECODE, NVL, LTRIM, RTRIM, SUBSTR,
> GREATEST, LEAST) in the [SQL
> reference|https://calcite.apache.org/docs/reference.html].
> Same goes for MySQL-specific functions (e.g. JSON_TYPE).
> I don't think we should have separate lists of Oracle-specific functions and
> MySQL-specific functions. Because quite a few functions appear in more than
> one place. Better, I think, to have a concise annotation against each
> function which tables it occurs in.
> The current list of tables is standard, oracle, spatial, mysql. Perhaps also
> indicate whether a function is an extension to the SQL standard but still
> occurs in Calcite's default table.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
